# AI Automation Project Brief

Use this blank brief to define one workflow before choosing a model, automation platform, or implementation partner. Replace each prompt with a concrete answer. If an answer is unknown, name the person responsible for resolving it and the date it is due.

Source and field guidance: https://ctbagents.com/resources/ai-automation-requirements-template/

## 1. Workflow and business case

- **Workflow name:**
- **Business owner:**
- **Problem and business impact:**
- **Trigger:**
- **Volume and frequency:**
- **Current steps and handoffs:**

## 2. Systems and data

- **Systems touched and source of truth:**
- **Inputs and required fields:**
- **Data sensitivity:**
- **Access constraints:**
- **Processing and retention constraints:**

## 3. Automation boundary

- **Desired output or end state:**
- **Where AI reasoning may help:**
- **Allowed actions:**
- **Prohibited actions:**
- **Human review and approval points:**

## 4. Exceptions and acceptance

- **Common exceptions:**
- **Required failure behavior:**
- **Representative normal cases:**
- **Representative edge cases:**
- **Representative failure cases:**
- **Objective acceptance tests:**

## 5. Measurement and ownership

- **Current baseline:**
- **Target result:**
- **Measurement window:**
- **Deployment owner:**
- **Monitoring and alerts owner:**
- **Exception reviewer:**
- **Post-launch support owner:**

## Pre-kickoff evidence

- [ ] Representative normal, edge, and failure-case inputs are available.
- [ ] API or integration access and permission scopes are confirmed.
- [ ] Consequential actions have written approval rules.
- [ ] A named operator owns production exceptions.
- [ ] Acceptance tests and the measurement window are agreed.
- [ ] Rollback, shutdown, and credential-revocation steps are documented.

Prepared from the free AI Automation Requirements Template published by Coding The Brains: https://ctbagents.com/resources/ai-automation-requirements-template/